Alvaro Herrera <alvherre@commandprompt.com> writes:
> Alvaro Herrera wrote:
>> Tom Lane wrote:
>>> BTW, shouldn't the log entry distinguish whether this was VACUUM,
>>> ANALYZE, or both?
>>
>> We don't actually log anything for ANALYZE (the logging code is in
>> lazy_vacuum_rel).
>>
>> Maybe it should be in autovacuum.c.

Actually, I had misunderstood where you were proposing to put this.
I believe that where you have it, the elapsed-time indication will
only cover the VACUUM step; so it's not relevant to this code whether
an ANALYZE would happen too.

My suggestion is that you add similar but independent logging to
analyze.c, controlled by the same min-duration variable. So the
log output would treat autovac and autoanalyze as two independently
loggable operations. I don't think there's much to print about
an autoanalyze except its runtime ... well, maybe you could print
the numbers of rows sampled and estimated, but I dunno if it matters.
The point of doing it is just to be able to track what the heck
autovacuum is doing ...

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 3: Have you checked our extensive FAQ?

http://www.postgresql.org/docs/faq

Tom Lane wrote:
> Alvaro Herrera <alvherre@commandprompt.com> writes:
> > Alvaro Herrera wrote:
> >> Tom Lane wrote:
> >>> BTW, shouldn't the log entry distinguish whether this was VACUUM,
> >>> ANALYZE, or both?
> >>
> >> We don't actually log anything for ANALYZE (the logging code is in
> >> lazy_vacuum_rel).
> >>
> >> Maybe it should be in autovacuum.c.
>
> Actually, I had misunderstood where you were proposing to put this.
> I believe that where you have it, the elapsed-time indication will
> only cover the VACUUM step; so it's not relevant to this code whether
> an ANALYZE would happen too.

Correct.

> My suggestion is that you add similar but independent logging to
> analyze.c, controlled by the same min-duration variable. So the
> log output would treat autovac and autoanalyze as two independently
> loggable operations. I don't think there's much to print about
> an autoanalyze except its runtime ... well, maybe you could print
> the numbers of rows sampled and estimated, but I dunno if it matters.
> The point of doing it is just to be able to track what the heck
> autovacuum is doing ...

Seems fair. Updated patch attached.

If there are no further comments, I'll add some docs and apply it later.

--
Alvaro Herrera http://www.CommandPrompt.com/
The PostgreSQL Company - Command Prompt, Inc.


---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?

http://archives.postgresql.org

Alvaro Herrera wrote:

> Seems fair. Updated patch attached.
>
> If there are no further comments, I'll add some docs and apply it later.

If autovacuum_vacuum_cost_delay is set to 20, my pitiful desktop system
takes 8.4 seconds to analyze a 8000 page table:

DEBUG: "bar": scanned 3000 of 8811 pages, containing 681000 live rows and 0 dead rows; 3000 rows in sample, 2000097 estimated total rows
LOG: automatic analyze of table "alvherre.public.bar" system usage: CPU 0.00s/0.01u sec elapsed 8.39 sec

So I guess logging on ANALYZE is not that removed from reality.

I also think we should remove these DEBUG messages, that are now
useless:

DEBUG: autovac: will ANALYZE bar

--
Alvaro Herrera http://www.CommandPrompt.com/
The PostgreSQL Company - Command Prompt, Inc.

---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to majordomo@postgresql.org so that your
message can get through to the mailing list cleanly

Alvaro Herrera <alvherre@commandprompt.com> writes:
> I also think we should remove these DEBUG messages, that are now
> useless:

> DEBUG: autovac: will ANALYZE bar

+1, those were just ad-hoc.

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster

On Tue, 2007-04-17 at 21:45 -0400, Alvaro Herrera wrote:
> Alvaro Herrera wrote:
>
> > Seems fair. Updated patch attached.
> >
> > If there are no further comments, I'll add some docs and apply it later.
>
> If autovacuum_vacuum_cost_delay is set to 20, my pitiful desktop system
> takes 8.4 seconds to analyze a 8000 page table:
>
> DEBUG: "bar": scanned 3000 of 8811 pages, containing 681000 live rows and 0 dead rows; 3000 rows in sample, 2000097 estimated total rows
> LOG: automatic analyze of table "alvherre.public.bar" system usage: CPU 0.00s/0.01u sec elapsed 8.39 sec
>
> So I guess logging on ANALYZE is not that removed from reality.
>
> I also think we should remove these DEBUG messages, that are now
> useless:
>
> DEBUG: autovac: will ANALYZE bar

Looks good.

--
Simon Riggs
EnterpriseDB http://www.enterprisedb.com



---------------------------(end of broadcast)---------------------------
TIP 6: explain analyze is your friend

I applied this patch some hours ago but I haven't gotten the
pgsql-committers mail and I don't see it in the archives either. Here
is the evidence:

http://developer.postgresql.org/cvsw...ml/config.sgml

Is there a problem with pgsql-committers again? Attached is the patch I
applied.

$ cvs commit -F msg
? msg
? cscope.out
? cscope.files
Checking in doc/src/sgml/config.sgml;
/cvsroot/pgsql/doc/src/sgml/config.sgml,v <-- config.sgml
new revision: 1.121; previous revision: 1.120
done
Checking in src/backend/commands/analyze.c;
/cvsroot/pgsql/src/backend/commands/analyze.c,v <-- analyze.c
new revision: 1.105; previous revision: 1.104
done
Checking in src/backend/commands/vacuumlazy.c;
/cvsroot/pgsql/src/backend/commands/vacuumlazy.c,v <-- vacuumlazy.c
new revision: 1.86; previous revision: 1.85
done
Checking in src/backend/postmaster/autovacuum.c;
/cvsroot/pgsql/src/backend/postmaster/autovacuum.c,v <-- autovacuum.c
new revision: 1.42; previous revision: 1.41
done
Checking in src/backend/utils/misc/guc.c;
/cvsroot/pgsql/src/backend/utils/misc/guc.c,v <-- guc.c
new revision: 1.386; previous revision: 1.385
done
Checking in src/backend/utils/misc/postgresql.conf.sample;
/cvsroot/pgsql/src/backend/utils/misc/postgresql.conf.sample,v <-- postgresql.conf.sample
new revision: 1.215; previous revision: 1.214
done
Checking in src/include/postmaster/autovacuum.h;
/cvsroot/pgsql/src/include/postmaster/autovacuum.h,v <-- autovacuum.h
new revision: 1.10; previous revision: 1.9
done
Collecting file lists...Done.
Sending mail to pgsql-committers@postgresql.org...Done

--
Alvaro Herrera http://www.CommandPrompt.com/
The PostgreSQL Company - Command Prompt, Inc.


---------------------------(end of broadcast)---------------------------
TIP 3: Have you checked our extensive FAQ?

http://www.postgresql.org/docs/faq